A PRAGMATIC RANDOMIZED TRIAL OF CYP2C19 GENOTYPING IMPLEMENTATION FOLLOWING PERCUTANEOUS CORONARY INTERVENTION (PCI) Sony Tuteja, PharmD, MS Twitter @sony_tuteja Perelman School of Medicine at the University of Pennsylvania, Philadelphia, PA

CYP2C19 polymorphisms and clopidogrel response Minor allele Frequency Europeans 0.15 African East Asian 0.29 South Asian 0.35 Trenk D et al. Thromb Haemost 2013 Scott et al. Clin Pharm Ther 2013

CYP2C19 polymorphisms and clopidogrel response Mega JL et al. JAMA 2010 Shuldiner AR et al. JAMA 2009

Conflicting recommendations exist for CYP2C19 testing Clinical Pharmacogenetics guidelines recommend prasugrel or ticagrelor in CYP2C19 LOF carriers. 1 FDA placed a “black box” warning on the clopidogrel label in 2010 recommending alternative agents among carriers of 2 LOF alleles. 2 ACC/AHA guidelines do not recommend routine CYP2C19 genetic testing. 3 1 Scott SA et al. Clinical Pharmacology and Therapeutics 2013 2 https://www.fda.gov/Drugs/DrugSafety/default.htm 3 Levine GN et al. Circulation 2016

Assessment of prospective CYP2C19 genotype guided Dosing of AntiPlatelet Therapy in Percutaneous Coronary Intervention (ADAPT) Rationale It is unknown how physicians will utilize the CYP2C19 test results in a real world setting Objective To provide evidence regarding the implementation and effectiveness of CYP2C19 testing Hypothesis Pharmacogenetic test results will influence prescribing of antiplatelet medications post PCI

Proportion of participants receiving prasugrel/ticagrelor The ADAPT study: A Pragmatic Randomized Clinical Trial PCI Usual care- no genotyping Genotyped Primary endpoint Proportion of participants receiving prasugrel/ticagrelor Secondary endpoints Agreement with the genotype guided antiplatelet recommendations Clinical Outcomes: Major Adverse Cardiac Events and Major Bleeding

Study Intervention Genotyped Group Usual Care Buccal swab for genotyping on the SpartanRx rapid turnaround device (CYP2C19 *2, *3, *17) Genotyped guided recommendations provided verbally CYP2C19 *2 or *3 carriers prasugrel or ticagrelor CYP2C19 *1 or *17 carriers  clopidogrel Antiplatelet choice remained at the discretion of the treating interventional cardiologist Usual Care Saliva collected for post study genetic analysis

Inclusion/Exclusion Criteria Age ≥ 18 and ≤ 80 years PCI with stent implantation Exclusion Need for imminent surgery History of intracranial hemorrhage, stroke Active bleeding Need for long-term anticoagulation Study staff unavailable to conduct randomization or genotyping

Sample size determination Rate of pre-study prasugrel/ticagrelor use (~20%) Anticipated increase in prasugrel/ticagrelor prescribing based on frequency of CYP2C19 LOF alleles (~30-35%) A sample size of 138 per group would provide 80% power at an alpha=0.05 to detect a 15% difference in prasugrel/ticagrelor prescribing between the groups Sample size was increased to 250 per group to allow for subgroup comparisons

Participant Demographics Genotyped N=249 Usual Care N=255 P-value Sex, n(%) Male 181 (73%) 190 (74%) 0.76 Race, n (%) White Black 194 (78%) 48 (19%) 197 (77%) 51 (20%) 0.99 Age 63.0 ± 9.7 62.9 ± 10.2 0.90 ACS 124 (50%) 129 (50%) 0.93

Participant History Genotyped N=249 Usual Care N=255 P-value Medical History, n(%) Hypertension Cholesterol PCI Diabetes MI CABG 190 (76) 112 (45) 83 (33) 89 (35) 63 (25) 32 (13) 199 (78) 113 (44) 79 (31) 67 (26) 36 (14) 0.67 0.80 0.63 0.26 0.84 0.70 P2Y12 prior, n (%) Clopidogrel Prasugrel Ticagrelor None 80 (32) 12 (5) 2 (1) 154 (62) 85 (33) 9 (4) 6 (2) 154 (60) 0.65

Procedure Characteristics Genotyped N=249 Usual Care N=255 p-value Length of stay, mean (SD), days 2.9 ± 3.7 3.1 ± 4.0 0.66 Drug eluting stents, n(%) 237 (95) 236 (93) 0.27 Number of stents, mean (SD) 1.3 ± 0.6 1.4 ± 0.6 0.73

CYP2C19 Genotypes for Intervention Group Genotyping results available 1.4 ± 0.4 hours post swab Genotype Frequency *1/*1 34% *1/*17 31% *17/*17 5% *1/*2 20% *2/*17 *2/*2 3% Inconclusive 4% 28%

Primary Outcome: Antiplatelet Drugs Prescribed Genotyped N=249 Usual Care N=255 P-value Clopidogrel Prasugrel or Ticagrelor 174 (70%) 75 (30%) 201 (79%) 54 (21%) 0.03 Fisher’s exact test

Prasugrel/ ticagrelor use greater in the LOF carriers Genotyped No-LOF LOF carriers N=174 n=68 Clopidogrel 136 (78%) 32 (47%) Prasugrel or Ticagrelor 38 (22%) 36 (53%) P<0.001 Usual Care N=255 201 (79%) 54 (21%) P<0.001

Agreement rate = LOF prasugrel/ticagrelor + non-LOF clopidogrel total number genotyped

Agreement with genotype guided recommendations CYP2C19 diplotype Phenotype Clop Pras/ Ticag *1/*2 *2/*17 *2/*2 Intermediate or poor metabolizer (n=68) 36 (53%) *1/*1 *1/*17 *17/*17 Normal or rapid metabolizer (n=174) 136 (78%) Non-agreement reasons 9- Stable CAD 6- Cost 3- Contraindications 3- MD preference 6- Disease characteristics 6- Patient already on therapy 5- ACS 5- Recurrent events 32 (47%) 38 (22%) Agreement rate 71% Non-agreement rate 29%

Prior antiplatelet therapy predicted antiplatelet drug choice independent of genotype Prior P2Y12 OR remaining on same 95%CI P-value Clopidogrel 2.04 1.22, 3.45 0.007 Prasugrel/ ticagrelor 99.3 13.2, 744 <0.0001

Genotype agreement rate Genotype did not influence prescribing among patient already on prasugrel or ticagrelor Prior P2Y12 Genotype agreement rate Clopidogrel (n=80) 76% Prasugrel/ticagrelor (n=14) 21% None (n=147) 73% P-value <0.0001

Clinical outcomes Genotyped (n=249) Usual Care (n=255) P-value Follow-up time (months) 17.2 (7.5) 16.1 (8.2) 0.14 MACE 34 (13.7) 26 (10.2) 0.27 BARC 3 or 5 bleed 6 (2.4) 8 (3.1) 1.0 MACE= myocardial infarction, stroke, death from cardiovascular cause, stent thrombosis, urgent revascularization BARC= Bleeding Academic Research Consortium

Summary CYP2C19 test results significantly influenced antiplatelet prescribing Genotype guided recommendations were followed 71% of the time Prior antiplatelet therapy significantly influenced the choice of antiplatelet drugs

Conclusions Antiplatelet prescribing was not universally in agreement with genotype suggested recommendations Physicians consider both clinical and genetic factors when prescribing antiplatelet agents following PCI

Acknowledgements ADAPT Study Team Cardiac Cath Labs Jay Giri Daniel J. Rader Henry Glick William Matthai Ashwin Nathan Karen Monono Craig Carcuffe Karen Maslowski Gene Chang Taisei Kobayashi Saif Anwaruddin John Hirshfeld Robert Wilensky Howard C. Herrmann Daniel M. Kolansky Cardiac Cath Labs Hospital of the University of Pennsylvania Penn Presbyterian Medical Center Department of Laboratory Medicine Funding Penn Center for Precision Medicine

Extra slide

Post-hoc analysis Non-LOF (n=326) LOF-clop (n=90) LOF-pras/ticag (n=52) MACE 33 (10.1) 14 (15.6) 9 (17.3) Major bleed 6 (1.8) 4 (4.4) 1(1.9) Non cardiac death 5 (1.5) 1 (1.1) 0 (0) Composite 42 (12.9) 19 (21.1) 10 (19.2) HR 1.84 95% CI 1.06 to 3.20 (p=0.03)